Queensland adult ECMO retrieval service: A description of the service and analysis of outcomes

Objective: To describe the Queensland Adult ECMO Retrieval Service (QAERS) and assess observed mortality of patients retrieved and treated with ECMO, against benchmarks. Design: Data was retrospectively collected from clinical and quality assurance databases at the three QAERS hospitals. Demographic data, diagnostic category, and hospital mortality were collected for patients referred to QAERS. Additional data was collected on patients receiving ECMO either before or after transport to a receiving hospital (ECMO patients), enabling calculation of RESP or SAVE scores. In ECMO patients with cardiogenic shock, individual risk of deaths were calculated by SAVE score. Monte Carlo analysis generated a discrete probability distribution function (PDF) of expected number of deaths, with 95 % confidence intervals (CI). The observed number of deaths was compared to this PDF. This was repeated for ECMO patients with respiratory failure, using RESP score. Setting: ICUs in Queensland and Northern NSW. Participants: All patients referred to QAERS from May 2017 to December 2023. Main outcome measures: Predicted and observed mortality of ECMO patients with cardiogenic shock or respiratory failure. Results: 237 patients were referred. 135 were retrieved, with 77 transported on ECMO. 11 commenced ECMO after transfer, giving a total of 88 ECMO patients. 35 ECMO patients had cardiogenic shock and 53 had respiratory failure. 16 cardiogenic shock patients died (95 % CI of PDF 17-28). 7 respiratory failure patients died (95 % CI of PDF 8-19). Conclusions: Observed mortality of patients retrieved and treated with ECMO was lower than mortality predicted by SAVE and RESP scores.